Thursday, December 3rd 2020
Intel Core i9-11900K "Rocket Lake" AotS Benchmark Numbers Surface
An alleged Ashes of the Singularity (AotS) benchmark results page for the top 11th Gen Core "Rocket Lake" processor leaked to the web courtesy TUM_APISAK. It's official now that Intel will keep its lengthy processor model number schemes, with the top part being the Core i9-11900K, a successor to the i9-10900K. It also confirms that the "Rocket Lake" silicon caps out at 8-core/16-thread, with performance on virtue of the IPC gains from the new "Cypress Cove" CPU cores."Cypress Cove" is believed to be a back-port of "Willow Cove" to the 14 nm silicon fabrication process that "Rocket Lake-S" is built on.
The screenshot also confirms the nominal clocks (base frequency) of the i9-11900K to be 3.50 GHz, as Intel tends to put base frequency in the name-string of its processors. Paired with a GeForce RTX 3080 and 32 GB of RAM, the i9-11900K-powered machine yielded 62.7 FPS CPU frame-rate at 1440p resolution, and 64.7 FPS CPU frame-rate at 1080p (a mere 3.18% drop in frame-rates from the increase in resolution). These numbers put the i9-11900K in the same league as the Ryzen 7 5800X in CPU frame-rates tested under similar conditions.
Sources:
TUM_APISAK (Twitter), 1440p Results, 1080p Results
The screenshot also confirms the nominal clocks (base frequency) of the i9-11900K to be 3.50 GHz, as Intel tends to put base frequency in the name-string of its processors. Paired with a GeForce RTX 3080 and 32 GB of RAM, the i9-11900K-powered machine yielded 62.7 FPS CPU frame-rate at 1440p resolution, and 64.7 FPS CPU frame-rate at 1080p (a mere 3.18% drop in frame-rates from the increase in resolution). These numbers put the i9-11900K in the same league as the Ryzen 7 5800X in CPU frame-rates tested under similar conditions.
52 Comments on Intel Core i9-11900K "Rocket Lake" AotS Benchmark Numbers Surface
The original ryzen shrank the gap, zen2 further to where it was mostly gaming and single thread performance where they lacked. Zen3 is the first to actually beat it in more things than not. Intel can't even get past 14nm and still has a very performant chip (price and power not withstanding). ;)
But, i see you are conveniently skipping on that part.
And if AMD decide to release, 5700X - 8c16t @ 65w, against the 11900k - 8c16t @ 125w, then what? Same point as above, although i find it amusing that you are celebrating the core count regression and spin it as some kind of win.
I've got mixed feelings on Rocket Lake, going backwards in core count is not a good look...
Again more cores is interesting if you do work with your PC and then again, when you need 8, you need 12, if you do work going for 8 is just silly.
5600X for gaming
5900X and higher for work
5800X if you are drunk and again, that is what this chip is trying to compete with….
You inferred we need more core/threads with the original post I quoted. Now, you post up how a 8c/16t processor works well.
Thank?! :)
Site needs a rumors section
I don't see a lot of point in 8 core either maybe it can do 5.5 on air lol but 8 core really is plenty for most people even potato console chips are 8 core so no matter what tubers think guess they haven't ever heard of virtual machines where more cores do matter, memory as well but I love it when people post click bait to ltt and gn what happened to 2cents guess he's silent on the matter lol :roll:
Sure gamers might not benefit from higher core count yet but there is no downside either. With 5900X/5950X high end users can have all the benefits of higher core count but none of the downside, well beside the cost of ownership.
With the 11900K Intel might force AMD to reduce the price of 5800X and that's it, nothing else change, 5900X and 5950X remain king. Kinda disappointing to see Intel going backward with their next gen chip on core count really.
Now, Intel squeezing more performance out of 14NM would be impressive if it didn't come at the expense of huge power draw and heat.
Now, I'm going to make a prediction that soon after Intel releases rocket lake, AMD will release a Zen3+/XT refresh and it will be on the improved 7nm EUV node. When Zen3 was released I wondered why it was on the same exact node as Zen2, but it makes sense as AMD didn't need the advantages of the improved 7nm node to take the crown and thus could keep costs the same and get more profit. Now, if Intel threatens with rocket lake, AMD can go to the improved 7nm EUV node, squeeze another 100-200Mhz, maybe even 300mhz, take the crown back, slap the XT suffix on the model number, and MOST importantly, justify keeping pricing exactly where it is.
The parity is achieved only through shear brute force, which results in immense power draw.
A better reason a gamer might want a 5900X or 5950X for gaming would be core binning and a second chiplet for a better chance at overclocking.
Intel 14nm +++: 24 by 24nm transistors
TSMC 7nm+: 22 by 22nm transistors
(he examined cache in particular).
I really cant see any reason to buy intel accept if your are a hardcore intel boy or need AVX 512. Else i think Zen 3 is the loggically way to go.
Of course AMD has a significant cost advantage for higher core-count CPUs using chiplet designs, but mainstream really doesn't need more than eight cores. The comments that Intel HEDT is dead is absolutely correct.
This is good for competition.
You can read up on this here. Really interesting article about porting a game (Detroit) that was originally planned to stay on PS4 to PC and the challenges they faced. Consoles are inherently better at multiscaling because draw calls are much faster, which mean you won't be bottlenecked by the render thread as much. The article also shows how you can't just assume that because consoles have 8 core CPUs this easily transfers to a PC environment and results in the good scaling in PC ports.
Also note that devs on consoles only have acces to 7 cores. Even if you were able to transfer multiscaling 1:1 from consoles to PC you'd still only scale with 7 cores max.
I think the 5800x is a waste of money for gaming. It's certainly not worth 150€ more than the 5600x, especially when for most of its lifespan the 5800x won't be much faster at all in games. That might change in 5+ years but is that really worth considering and spending an extra 150€ on? I really don't think it is.
I've been saying this since Pentium days .. if it wasn't for AMD to this day we would have be still rocking Pentiums 4.
If somebody would just go over my comments on this forum over the years it would look like I'm AMD fanboy but in reality I always cheer for the underdog for sake of competition.
Hell ... thanks to AMD I just bought only my second brand new Intel cpu i7 9700k.
I've owned plenty Intel CPUs but only getting a good deal buying them used because looking at the prices for new AMD always had the better price/perf in my price range over the years.
Intel doesn't make it easy to be a fan ... I found out that i7 9700k doesn't have HT only after I installed the CPU. One gen they remove HT on i7 next gen they decide to add it to i5. Fucking Intel! :D It all depends on the system and what you are using it for really.
All those test to prove that you need more CPU cores are using $1000+ GPUs where in reality in most scenarios the GPU will be the bottleneck.
For example I upgraded from i5 6600k to 9700k and I honestly haven't noticed any difference because I'm using GTX 1070 which eventually will be upgraded to 3060ti where I expect 9700k to come into play.
Also it would be really silly people to claim that Intel still pushing 8c because it's the "sweet spot" it's, we all know what's the story there.
Also, they are not able to keep up with core count due to insufficient die space (as the node is not dense enough). The core count regressed from Comet Lake to Rocket Lake purely because the latter will use up more transistors, thus, Intel have to reduce the core count to fit within the die size budget. Not to mentioned they integrated the Xe graphics in it, which itself takes up a sizeable space on the die.